{"id":"W3027597167","doi":"","title":"Daydream: Accurately Estimating the Efficacy of Performance Optimizations for DNN Training","year":2020,"lang":"en","type":"article","venue":"USENIX Annual Technical Conference","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Profiling (computer programming); Artificial neural network; Machine learning; Software; Artificial intelligence; Graph; Parallel computing; Computer engineering; Theoretical computer science; Programming language","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001647722,0.0001633516,0.0002314061,0.00002908962,0.0002812232,0.00007192502,0.001595113,0.00006976046,0.000007022008],"category_scores_gemma":[0.0008088627,0.0001257725,0.00007910027,0.0006929539,0.0001798663,0.0004848946,0.0003519785,0.0002541055,0.00000792932],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001495348,"about_ca_system_score_gemma":0.0001364111,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00000122225,"about_ca_topic_score_gemma":0.000001464714,"domain_scores_codex":[0.9985968,0.00003586758,0.0004255599,0.0004293023,0.0002114768,0.0003009796],"domain_scores_gemma":[0.9977794,0.001069811,0.0002192998,0.0005353065,0.0002717982,0.0001243576],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00006467305,0.0001396194,0.00008837552,0.00007484999,0.00003244319,0.000002101736,0.005360084,0.1899517,0.006124929,0.4492358,0.001795037,0.3471304],"study_design_scores_gemma":[0.0003504488,0.000208944,0.0004514196,0.00003942054,0.00001566814,0.000009317775,0.00007406242,0.9899954,0.002565103,0.001950754,0.004133523,0.0002059634],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.004669628,0.00002149988,0.9875934,0.006147426,0.00005597109,0.0006197451,0.00002856107,0.0003568895,0.0005069318],"genre_scores_gemma":[0.6421368,0.00000656749,0.3573345,0.0003307179,0.00006922004,0.00009310098,0.000006010052,0.000009192996,0.00001389909],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8000436,"threshold_uncertainty_score":0.5128852,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1356333107934018,"score_gpt":0.3383242010037694,"score_spread":0.2026908902103675,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}